Unique maps limit

Unique maps limit

Return to topic list

Wed Aug 11 14:44:31 2021   by   Mizuki
Does the fact that there are only 100000000 seeds mean this program can only generate 100000000 unique maps?
 
Wed Aug 11 17:21:52 2021   by   Torben
It is only the web version that is limited to 100000000 seeds.  The command-line version can take many more different seeds.

Additionally, there are other parameters than the seed that can change the look of a planet.  A simple example is the water level.  The command-line version has many other options that can modify the look of a planet.

But do you really need more than a hundred million different planets?
 
Sun Sep 12 02:02:45 2021   by   Avarus Lux
while using the offline version (update: June 23 2021) i found that starting from seed input 00000 (format: 00000, 00001, 00002, 00003, etc,) using any seed with more then 5 digits and above the number 46337 will not work for some reason.

seed input 0 or 00 or 00000 and equally 3 or 03 or 003 or 00003 gives the same seed results so the range seems to be 00000 to 46337 for seed options.

Utilizing the following command parameters in windows 10 PowerShell:
C:\Users\LUX\Desktop\Project\planet\planet.exe -pm -E10 -E-4 -s 02245 -b -w 3000 -h 3000 -C C:\Users\LUX\Desktop\Project\planet\OlssonLight.col -o C:\Users\LUX\Desktop\Project\Maps\SeedFresh02245.bmp

Changing the water level gives me some variation within these 46337 seeds but is not quite close to a hundred million different seed options unless i am missing a key instruction to go beyond that what i have been able to currently generate?

i already found this amount of seeds suspiciously low so i expect something is not quite right and i am wondering what i might be doing wrong or if there is something else going on.
 
Mon Sep 13 11:02:47 2021   by   Torben
The offline map generator can use non-integral seed numbers, e.g. 7.236481.  In fact, the integral numbers used by the online version are converted to fractional by adding "0." in front of the number, so 6578 becomes 0.6578.
 
Sun Sep 19 17:54:12 2021   by   Avarus Lux
Thank you for that non-integral number instruction... that extends the range from just those 5 integral digits up to 5 digits before the decimal point with up to 15 digits after the decimal point (15+/more digits are ignored) which gives near limitless unique results.
 
Thank you Torben.
 
Mon Sep 20 10:43:31 2021   by   Torben
One thing to note: If two seeds differ only in the lowest precision digits, the maps will have similar overall structure but will differ in details.  This can be used to make small variations on a map.
 
Sat Oct 30 09:27:42 2021   by   Mizuki
What's the limit on the offline version? Lux seemed to imply that it's 15 digits after decimal.
"differ only in the lowest precision digits"
Is it digit or digits? If the latter, then what is the crossover between "similar overall structure" and completely different seed? e.g 0.55550/1 vs 0.555555555555550/1
 
Mon Nov 1 11:04:22 2021   by   Torben
Acceptable seeds for the offline version are from ca. -46344.0 to 46337.8 -- numbers outside this range will cause errors.  I usually use seeds between 0 and 1.

As for when maps start to be different, it is a gradual thing.  For seeds between 0 and 1, maps will be quite different if seeds differ in the 8th digit after the decimal point, but after that they will be gradually more alike.  With difference in the 9th digit, the overall placement of continents will be similar, from the 10th digit, the overall shape of continents will also be similar, and the similarities increase until you need zoom to see the differences when the seed differs only in the 15th digit after the decimal point.  After that, you hit the precision of floating point numbers, so you won't see any differences.
 
Wed Nov 3 17:32:47 2021   by   Mizuki
planet -s 0.1465465464464134165415432745241541324324654357454174654987634196784324657452875 -o 1.bmp

planet -s 0.1465465464464134165415432745241541324324654357454174654987634196784324657452875 -o 2.bmp

certutil -hashfile 1.bmp sha1
SHA1 hash of 1.bmp:
10d31f98ccb0d39a79e06e207804666b46c72e40
CertUtil: -hashfile command completed successfully.

certutil -hashfile 2.bmp sha1
SHA1 hash of 2.bmp:
5cf2b5485ea26059483c8b9cad0a855d1c1bc79c
CertUtil: -hashfile command completed successfully.

yeah, as i expected lol
I wonder where the difference come from, a subtle change of some of the pixels back and forth (e.g 32555E vs 32555F)?
 
Wed Nov 3 17:37:55 2021   by   Mizuki
also, that would suggest the limit on the number of unique maps possible is somewhere around 9x1019
uhhhhh
that should be enough i guess lol
 
Wed Nov 3 18:53:12 2021   by   Mizuki
also, that would suggest the limit on the number of unique maps possible is somewhere around 9x1019
uhhhhh
that should be enough i guess lol
 
Wed Nov 3 18:53:20 2021   by   Mizuki
also, that would suggest the limit on the number of unique maps possible is somewhere around 9x1019
uhhhhh
that should be enough i guess lol
 
Thu Nov 4 11:51:53 2021   by   Torben
The different SHA1 hashes are because the command line is appended to the bmp file (for seed recovery etc), and since you use different output files, this affects the hash code.
 

Return to topic list



New message:
Topic:
Posted by:

Type the values of the dice shown below:

Return to topic list